Several potential advantages have been ascribed to the use of phakic intraocular lenses (IOLs) for the correction of high myopia and hyperopia, including excellent refractive accuracy, preservation of accommodation, procedural compatibility with established methods for aphakic IOL implantation, and reversibility.1 Although an anterior chamber IOL is especially attractive because of the ease at which it can be inserted into the eye and fixed in the iridocorneal angle, the first generation phakic IOL of this genre (Domilens Model ZB) was associated with an unacceptably high incidence of endothelial changes characterized by a reduction in cell density, marked pleomorphism, and the appearance of dark zones and areas of acellularity.2"4 Since these changes were presumed to be caused by excessive contact between the edge of the lens optic and adjacent endothelium- separated by only 0.71 to 1.50 mm in one study3- the ZB Model was redesigned as the Chiron/Domilens Model ZB5M to increase this distance by approximately 0.6 mm, in addition to making the loops more flexible and reducing the thickness of the optic at its center and periphery.
The purpose of the current report is to describe results during 3 years follow-up in a projected 5year study of high myopic patients implanted with the second generation Baikoff anterior chamber IOL Model ZB5M, with particular emphasis on the safety data that relate to lens-induced changes the in corneal endothelium.
As a result of this study, further modifications have been made to create a third generation Baikoff anterior chamber IOL to further reduce complications. The Baush & Lomb Surgical/Chiron Vision Model NuVita MA20 employs a redesigned optic, loop, and footplate to eliminate refracted glare, to reduce pressure in the angle that may lead to pupil ovalization, and to insure clearance of the iris and natural lens regardless of refractive power.
PATIENTS AND METHODS
After first obtaining written informed consent, five patients meeting the entry requirements of the study were enrolled at each of the nine participating centers, all of which were in France. Patients were given a complete ophthalmologic exam, implanted in one eye with the ZB5M lens (Baush & Lomb Surgical/Chiron Vision, Rochester, NY), and reexamined at 1 day, 1 week, and 3 months after treatment. After analyzing the 3 month specular microscopy data to assure that the lens was welltolerated, each center was asked to enroll an additional ten eyes. Patient examinations were also scheduled for 6, 12, 18, 24, 36, 48, and 60 months following lens implantation.
The Institutional Review Board-approved study protocol specified that patients had to be 25 to 45 years of age, and have a spherical equivalent and manifest refractive error between -8.00 and -15.00 diopters (D), at least 2500 cells per mm2 endothelial density (center/peripheral average), and an anterior chamber depth of 3 mm or more. Exclusion criteria included the presence of any of the following ocular disorders: inflammation of the anterior or posterior segment, chronic keratitis, corneal dystrophy, iris atrophy or rubeosis, aniridia, cataracts, history of any type of vitreous pathology, retinal disease (detachment, macular degeneration, diabetic retinopathy), microphthalmia, and/or glaucoma.
Model ZB5M IOL
The ZB5M Model is a single-piece, biconcave anterior chamber lens designed by George Baikoff, and based upon a multiflex Kelman anterior chamber IOL.5 It is manufactured by Baush & Lomb Surgical/Chiron Vision, Claremont, California (formerly Chiron Vision France/Domilens) from polymethylmethacrylate (PMMA) containing an ultraviolet blocker. The lens comes in overall lengths of 12.5, 13.0, and 13.5 mm, is angulated posteriorly at 20°, and the optic is 5 mm in diameter (effective optical diameter 4 mm). It is available in one diopter increments from -7 to -20 D, and is provided in ethylene oxide sterilized packaging.
where A = the patient's far point position in mm; E = virtual object distance for the 1OL in mm; F = IOL image distance in mm; DI = implant power in diopters; R = spectacle correction in diopters; Dc = power of the cornea in diopters; d = anterior chamber axial depth in mm; k = 1 mm and is the distance from the IOL to the crystalline lens; 12 = the distance from the spectacle plane to the cornea in mm; 1000 = a conversion factor for meters to mm; 1336 = the aqueous humor index of 1.336 multiplied by the mm conversion factor.
Although the procedure used to implant the Model ZB5M lens was not standardized for this particular study, the basic elements of the method employed by each of the nine centers is described. The patients were first given either general or local anesthesia followed by instillation of pilocarpine drops to induce miosis and help prevent damage to the anterior crystalline lens capsule during implantation. A 6 to 7 mm penetrating incision was made in the lateral nasal or temporal cornea parallel to the iris plane, allowing entrance into the anterior chamber which was filled with a viscoelastic before inserting (with forceps) the leading haptic of the Model ZB5M lens into the iridocorneal angle along the horizontal axis. Once the leading haptic was properly positioned, additional viscoelastic was added as needed, and the trailing haptic was then placed in the iridocorneal angle on the opposite side of the eye. After inspecting to see that the pupil was round and the lens had not rotated, the surgical incision was closed with four to five interrupted nylon sutures prior to examining the iridocorneal angle with a gonioscope. Care was always taken to remove viscoelastic before wound closure to avoid postoperative ocular hypertension. Postoperative treatment consisted of patching the eye for 48 hours, and topical instillation of a corticosteroid/antibiotic combination for up to 6 weeks. Nonsteroidal antiinflammatory agents were also used in some centers, and selected eyes were treated with an antiglaucoma medication.
Iridectomies were performed on patients presumed to be at risk of developing pupillary block. It was advised, however, that the resection not be excessively peripheral to avoid footplate entanglement should the lens become slightly displaced.
Patients were scheduled for evaluation preoperatively, during the immediate postoperative period (day 1, week 1), and at various intervals thereafter over the succeeding 5 years. Procedures performed included endothelial specular microscopy (central and peripheral), assessment of refraction (subjective/cycloplegic), visual acuity (distance, Monoyer scale; near vision, Parinaud scale), and intraocular pressure, gonioscopy, slit-lamp microscopy, fundoscopy, and patient self-evaluation of visual comfort.
Specular microscopy was performed at each center using the procedure in effect at the center and not a standard protocol. Although the methodology varied somewhat from center to center, the basic procedure employed in all centers follows. Endothelial cells were counted using a grid that was precalibrated knowing the dimensions of the sides of the 0.1 ? 0.1 mm square and the cone magnification of the microscope. Depending on the size of the field of the microscope, either 25 to 45 cells were counted in a small square (Cooper Vision specular microscope CEM4, three centers), or 100 to 130 cells were counted in four squares (Konan Keeler Pocklington specular microscope SpI, four centers; Konan clinical specular microscope 580, one center). Cell density per mmp 2 (ED) was calculated from the formula: ED = (n x 100) divided by C, where n is the number of cells counted per square, 100 is the mmp 2 conversion factor, and C is the number of evaluable squares.
Slit-lamp microscopy was performed to assess the effect of the Model ZB5M lens on various ocular structures, but particularly the iris, which is susceptible to deformation along the axis of the IOL haptic. The extent of IOL-induced iris retraction with pupillary ovalization was not measured objectively, but was subjectively evaluated relative to length, clinical progress, and ultimately the presence of goniosynechia which could be detected by the investigator during gonioscopy. Two forms of pupillary ovalization were noted. Discrete ovalization was defined as a slight elliptical deformity of the pupil in the axis of the IOL haptic that was stable over time or spontaneously reversible, and not always evident under slit-lamp examination. Progressive ovalization was an elliptical deformity that increased over time and could be irreversible, necessitating surgery if goniosynechia or synechia didn't allow iris contraction or miosis.
The investigator was also responsible for evaluating the patient's visual comfort and perception of halos and glare. Visual comfort was scored as very good, good, moderate, or bad. Patients were asked to report the presence or absence of halos (perception at night of a circular band of colored light around a light source such as the moon) and/or glare (intense, blinding light) during the early postoperative examinations (3, 6, 12, and 18 months).
Descriptive statistics were generated for all variables. Additionally, because endothelial cell density was considered to be the most critical study variable, it was further analyzed using the Generalized Estimating Equations methods described by Biggie et al7 which adjusts for the correlation between repeated measurements on the same patient. For this analysis, patient total endothelial cell counts at each interval after implantation were expressed as a proportion of their baseline value and fitted to the following equation which assumes cell loss to be constant over time: cells = e(a + bxt), where a is the y-intercept which corresponds to early cell loss presumably due to surgery, b is the rate of cell loss, and t is the time in years. Log transformation of cell count proportions (log cells = a + bxt) allowed for the use of ordinary linear regression for data modeling, and gave results that were similar to those obtained with nontransformed data. In addition to linear regression, multivariate analyses were also performed, with implant power, anterior chamber depth, contact lens usage (yes/no), and each surgical center being added to the model as factors, along with time after surgery. Repeated measures analysis of variance (ANOVA) was also used in a subset analysis of spherical equivalent refraction.
Figure 1: Mean (range) spherical equivalent manifest refraction over 3 years after implantation of the Baikoff ZB5M phakic IOL. Proportion of eyes within ±0.50. ±1.00, and ±2.00 D of emmetropia are underlined.
One hundred and twenty-one patients (134 eyes) were enrolled in the protocol, beginning December 11, 1990 and ending October 27, 1993. One patient was excluded because their age (62 years) fell well-outside the demographic range for this procedure, leaving 120 patients (133 eyes) for evaluation. The average age of the évaluable patients was 33.6 ± 6.3 years (range, 20.4 to 45.3 yr), 55.5% were female, 62.2% wore contact lenses at entry, and the average spherical equivalent refraction of the treated eye at baseline was -12.50 ± 2.60 D (range, -7.00 to -18.80 D). One hundred and seven of these patients (89.2%) were implanted with the Model ZB5M lens in one eye (right eye in 54.6%) and 13 (10.8%) in both eyes. The average anterior chamber depth of évaluable eyes was 3.5 ± 0.4 mm (range, 3.0 to 4.5 mm). Mean follow-up for the 133 eyes was 35.8 ± 9.6 months (range, 17.7 to 52.2 mo), with the actual number of eyes evaluated at each postoperative examination as follows: 6 months (104 eyes), 1 year (91 eyes), 18 months (78 eyes), 2 years (68 eyes), and 3 years (35 eyes).
The average power of the Model ZB5M lens implanted in this study was -11.80 ± 2.10 D (range, -7.00 to -15.00 D), with the majority (88.5%) being above -9.00 D. Fifty and eight-tenths percent of the lenses had an overall length of 13 mm, 35.4% were 12.5 mm, 13.8%> were 13.5 mm, and 49 of 90 (54%) exceeded the horizontal whiteto-white diameter by >0.5 mm; seven of 90 (7.8%) exceeded it by >1.0 mm. Nearly all of the investigators (88.9%) used Healon brand viscoelastic to facilitate lens implantation, and three investigators performed iridectomies on one, four, and seven eyes.
Routine postoperative management at eight of the centers included instillation of dexamethasone/neomycin (the ninth center used dexamethasone/chloramphenicol), and three used topical indomethacin or flurbiprofen.
Figure 2: Mean (range) spherical equivalent manifest refraction in 21 implanted eyes examined at all intervals. Proportion of eyes within ±0.50. ±1.00, and ±2.00 D of emmetropia are underlined. Values are similar to those in the entire population in Figure 1.
Figure 3: Refractive correction error is presented as the difference between achieved and intended refractive outcome. Means (ranges) are provided, and the proportion of eyes within ±0.50. ±1.00, and ±2.00 D of emmetropia are underlined.
Refraction- Spherical equivalent refraction for all eyes implanted with the Model ZB5M lens averaged approximately -1.00 D over the first 3 years of followup, and none were overcorrected by more than +1.00 D (Fig 1). Refractive stability during the first 2 years after implantation was good, with 32% to 41% of the eyes having a residual refractive error within ±0.50 D, 62% to 66% within ±1.00 D of emmetropia, and 87% to 91% within ±2.00 D, declining to 17.1%, 51.4%, and 80.0%, respectively, at the 3-year examination.
Stability of spherical equivalent refraction was also evaluated in a subset (21 eyes) of the total patient population for which follow-up information was available at each examination (Fig 2). Data obtained for the 21 eyes that met this criterion was similar to those presented in Figure 1 on the total number of implanted eyes. Analysis of the subset data using repeated measures analysis of variance revealed that the effect of time on spherical equivalent refraction was not statistically significant (p = 0.0139). Moreover, the average change in spherical equivalent refraction between 6 months and 3 years for the 21 eyes was -0.37 D (eyes became more myopic), the range was -1.50 to +1.25 D, and the confidence interval was -0.6714 to +0.0614, with 76.2% (16 eyes) within ±1.00 D of the 6 month refraction at 3 years.
Figure 4: Mean (range) uncorrected distance visual acuity (Snellen/decimal) over 3 years after implantation of the Baikoff ZB5M phakic IOL Proportion of eyes with visual acuity of 1 .0, 0.8, and 0.5 or better are underlined.
Figure 5: Mean (range) spectacle-corrected distance visual acuity (Snellen/decimal) over 3 years after implantation of the Baikoff ZB5M phakic IOL. Proportion of eyes with visual acuity of 1 .0, 0.8, and 0.5 or better are underlined.
Figure 6: Mean (range) uncorrected near visual acuity (Snellen distance equivalent/decimal) over 3 years after implantation of the Baikoff ZB5M phakic IOL.
Figure 7: Mean (range) spectacle-corrected near visual acuity (Snellen distance equivalent/decimal) over 3 years after implantation of the Baikoff ZB5M phakic IOL.
The mean error in refractive correction, as measured by the mean difference between achieved and intended correction8, was substantially less than 1.00 D over the entire study (Fig 3). At 3 years follow-up, 88.2% (117 eyes) were still within ±2.00 D of intended correction, 58.8% (78 eyes) were within ±1.00 D, 32.4% (43 eyes) were within ±0.50 D, and the maximum overcorrection was +1.30 D.
Distance Visual Acuity- As shown in Figure 4, uncorrected distance visual acuity averaged 0.048 at baseline, improving to 0.50 to 0.52 over the succeeding 2 years, and then declining somewhat to 0.45. The proportion of eyes with postoperative uncorrected visual acuity scores was: >0.5, 37.8% to 57.1%; >0.8, 13.5% to 21.5%; and >1.0, 3.7% to 7.2%, with the lowest percentages being recorded at the 3 year follow-up examination for 0.5 and 0.8 but not 1.0.
Spectacle-corrected visual acuity improved from a mean 0.54 before implantation to approximately 0.7 after, accompanied by an increase in the proportion of eyes with scores >0.5 and >0.8 (Fig 5). Overall, 36.2% to 50.7% of the eyes at the various postoperative examinations gained two or more Snellen lines of spectaclecorrected visual acuity (lowest percentage 3 years); 2.1% to 8.3% lost two lines (highest percentage at 3 years). Only one eye (2.8%) lost three lines of spectacledistance visual acuity (3 year examination only), and none lost more than three lines.
Endothelial Cell Density in Eyes Implanted with the ZB5M Phakic IOL*
Near Visual Acuity- The improvement in uncorrected near visual acuity paralleled that seen with distance vision, increasing from 0.21 at baseline to 0.68 to 0.74 after implantation (Fig 6). Spectacle-corrected near visual acuity also improved, but not nearly as much as its distance counterpart (Fig 7).
Endothelial Cell Density- Reduction in mean endothelial cell density for both the central and peripheral cornea in the implanted eyes averaged 3.3% at 6 months, declining by an additional 1% to 2% over the remaining 2.5 years of follow-up (Table 1). Although the proportion of eyes with endothelial cell loss in excess of 15% at any postoperative visit was as high as 14.8% (four of 27 eyes at 3 years), absolute counts over either region of the corneal endothelium rarely fell below 2000 cells/mmp 2 (one eye at 3 months declined to 1200 cells/mmp 2, one eye at 18 months to 1969 cells/mmp 2, and one eye at 24 months to 1800 cell/mmp 2).
Figure 8: Percentage of eyes with iris retraction and pupillary ovalization over 3 years after implantation of the Baikoff ZB5M phakic IOL
Figure 9B: Two- and three-dimensional finite element analysis were utilized to ensure the modifications would not compromise the safety data obtained from the multicenter study reported in this article. A two-dimensional simulation ensured that the same level of compression force was applied to each footplate and a three-dimensional simulation kept the axial displacement of the IOL unchanged.
Single factor linear regression of log transformed endothelial cell density data (expressed as a proportion of baseline values) yielded a y-intercept derived value of 3.8% (95% confidence interval, 2.3 to 5.2; p < 0.001) and an estimate of cell loss over time of 0.7% per year (95% confidence interval, +0.1 to -1.5; p=0.091), suggesting that most (3.8%) of the observed reduction in endothelial cells over the course of the study was not from the postoperative effects of the Model ZB5M IOL, but was attributable to the acute effects of surgery. The confidence interval around the estimate for nonsurgery-related cell loss over time predicts that endothelial cell density would not be expected to decline by more than 1.5% per year within the first 3 years of lens implantation, and could even increase by 0.1% per year (ie, remain unchanged).
Figure 9A: Modifications to the second generation Baikoff ZB5MF were made on the basis of clinical observations, leading to a third generation MA20 NuVita IOL. The incidence of glare and halos led to a redesign of the optic. The incidence of pupil ovalization resulted in loop and footplate design modifications.
Results of the multivariate analyses failed to reveal a statistically significant effect for any of the variables included in the model except that two centers had significantly greater endothelial cell loss than the other seven (eg, at 1 year after implantation, endothelial cell density was reduced by a mean 7.5 ± 1.4% at centers 5 and 6 in contrast to a 3.1 ± 1.7% reduction at centers 1, 2, 3, 4, 7, 8 and 9). Interestingly, the regression coefficient derived value for time in the multivariate model was 0.5% per year (95% confidence interval, +0.3 to -1.2; p = 0.173), which is quite similar to the estimate obtained from single factor linear regression (0.7% per year).
Halos/Glare- Halos, glare, or similar phenomena were reported by patients in 37 of 133 eyes (27.8%), beginning with 9.2% (11 of 120 eyes) of those followed at 1 month, increasing to 18.8%) (19 of 101 eyes) at 1 year, then declining to 12.5% (five of 40 eyes) at 3 years.
Iris Retraction- Iris retraction with pupillary ovalization was observed in 30 of 133 eyes (22.6%), with 5.8% (seven of 120 eyes) exhibiting this at 1 month, 9.9% (10 of 101 eyes) at 1 year, 16.7% (12 of 72 eyes) at 2 years, and 27.5% (11 of 40 eyes) at 3 years (Fig 8). Pupillary ovalization was characterized as discrete with no clinical significance in 27 of the eyes, while in two eyes it was considered progressive and in another eye it was associated with iris abnormalities.
Figure 1OA: To ensure constant clearance of the natural crystalline lens regardless of refractive power, the ZB5MF biconvex optic (left) was changed to the MA20 meniscus design with a fixed posterior curvature (right). The central optic zone was increased from 4.0 to 4.5 mm while the physical optic diameter remained 5.0 mm.
Figure 10B: Glare reduction was addressed by minimizing optic edge thickness approximately 20% and applying peripheral detail technology, a non-glare edge treatment.
Figure 10C: The loop and footplate profile was changed from a step vault (left) to a curved angulation (right) to eliminate contact with the iris. The central vault distance remains constant at 1 .00 mm, identical to the ZB5MF.
Figure 10D: The footplates were modified to better conform to angle geometry and reduce compression forces on the angle. The connecting bridge was positioned further away from the angle.
Intraocular Pressure- Intraocular pressure was increased above 20 mmHg in 8.2% (eight of 97 eyes) examined at day 2, 7.8% (eight of 102 eyes) at day 10, and 6.9% (eight of 115 eyes) at day 30, but in only 1.1% (one of 90 eyes) at 12 months and 0% (zero of 37 eyes) at 36 months. Overall, early transient elevation of intraocular pressure (^30 days) was present in 23 of 133 eyes (17.3%).
Astigmatism- Mean refractive astigmatism was 1.10±1.10D (range, 0 to 4.50 D) preoperati very, and was never above 1.10 D at any subsequent examination, with maximum astigmatism varying from 2.50 to 4.00 D. Increases in astigmatism in excess of 1.00 D were seen in approximately 15% of the eyes at any examination. At 1 month follow-up, at least one suture was removed in 12 of 117 eyes (10.3%) and at 3 months follow-up, at least one suture was removed in 15 of 117 eyes (12.8%).
Other Complications- A small number of eyes (=s3.8%) presented with corneal edema, uveitis, pupillary block, or a flat anterior chamber (Table 2). 10L displacement occurred in two of 133 eyes (1.5%) and rotation in six of 133 eyes (4.5%), three of which required removal and replacement with a second lens. A fourth eye also needed an IOL exchange because of a loop foot in the iridectomy, and in three of the eyes IOLs were removed but not replaced. A single retinal detachment occurred in an implanted eye in this series of patients, but only at a site that had been photocoagulated preoperatively. Once an IOL was removed, the patient was discharged from the study, so no follow-up information on these eyes is available.
Complications in 134 Eyes Implanted with the ZB5M IOL
Patient Visual Comfort- At 6 months, visual comfort was judged by 98.2%> of patients to be good or very good, at 12 months, 97.9%, and at 18 months, 93.9% of patients judged comfort to be either good or very good.
The results described were derived from an interim analysis of a study that was designed to provide 5 year follow-up data on phakic patients with high myopia implanted with the Chiron/Domilens Model ZB5M second generation Baikoff anterior chamber angle-fixated phakic IOL. The primary purpose of this study was to determine if the design changes that had been incorporated into the Model ZB5M lens were sufficient to reduce or eliminate the damage to the corneal endothelium that was associated with the first generation Model ZB phakic IOL.2'4 It is apparent from the data in Table 1 that most of the endothelial cell loss observed during the first 3 years following implantation of the Model ZB 5M lens was present at the 6 month postoperative examination, with only a mean 1% to 2% reduction occurring thereafter accompanied by an occasional decline (2.2% of all eyes) in individual counts below 2000 cells/mm2. Estimates from single factor regression analysis of these data suggest that 3.8% of the endothelial cells were actually lost at the time of surgery or shortly thereafter, and any decrease beyond this initial surgery-related decrement was quite gradual (0.69% per yr). The 0.69% per year endothelial cell loss estimate for eyes implanted with the Model ZB5M lens is slightly greater than the 0.33% to 0.53% per year estimates obtained for normal eyes in patients of comparable age who were subjected to specular microscopic evaluation in cross sectional studies.912 This estimate is, however, still well below the 1.5% per year reduction postulated as producing corneal decompensation in a 95 year old patient implanted with a phakic IOL at 25 years of age who experienced a 12% cell loss at the time of implantation.13
The only study variable found in a multivariate analysis to be significantly related to reduction in endothelial density was the effect of surgical center; two of the centers experienced somewhat greater cell loss than the other center. This finding is important because it suggests that consistent clinical outcome with regard to the magnitude of endothelial cell damage in eyes implanted with the Model ZB5M lens can be achieved in many different centers, although some will have better results than others. Indeed, one report of a study in which 28 eyes of 17 patients were implanted with the Model ZB5M lens by the same surgeon and followed for 2 years demonstrated that endothelial cell loss increased from a mean 7.5% at 3 months to 12.3% at 12 months, and then remained unchanged during the final year of follow-up.14 The y-intercept and regression coefficient derived from the data in that series using the equation in our report were 10.9% and 0.87% per year, indicating that nearly three times as many endothelial cells were lost as a consequence of surgery in comparison to the current study, but the overall rate of cell loss was similar.
Although the principal safety variable for this study was specular microscopy-determined change in corneal endothelial cell density over time, additional parameters were also evaluated to allow a more complete assessment of the nature and incidence of other adverse events. Patients reported that 37 of the 133 implanted eyes (27.8%) had halos, glare, or similar phenomena at some time over the course of the study; approximately 15% at each follow-up examination. Iris retraction with pupillary ovalization was generally of no clinical significance; it was observed in a total of 30 of 133 eyes (22.6%), and increased in incidence from 9.9% (10 of 101 eyes) at 1 year to 16.7% (12 of 72 eyes) at 2 years and 27.5% (11 of 40 eyes) at 3 years (Fig 8). Other complications, most of which resolved within 30 days of treatment, included corneal edema, uveitis, and pupillary block (Table 2). Implant displacement occurred in two of 133 eyes (1.5%) and rotation in six of 133 eyes (4.5%), necessitating exchanges in three of 133 eyes (2.3%), along with another eye (0.8%) in which the loop of the Model ZB5M lens passed through an iridectomy. Implants also had to be removed (without replacement) in an additional three eyes, one due to halos, and the other two, severe inflammation associated with a flat anterior chamber. Intraocular pressure and astigmatism were not affected adversely by the presence of the Model ZB5M lens in the anterior chamber.
Retinal detachment occurred in one eye, originating at a site which had been photocoagulated prior to implantation. The absence of retinal detachments is particularly noteworthy given that they have been previously reported in patients implanted with ZB and ZB5M lenses in some series15·16 (but not all17), and one group of investigators15 have opined that retinal detachment, probably caused by intraoperative hypotony, should be considered "a potentially important hazard in this type of refractive surgery."
One of the putative advantages of phakic IOL implantation over corneal surgery for the correction of high myopia is that an IOL is inherently more accurate than reshaping the cornea. The variable in this study that provides the most appropriate assessment of the accuracy or predictability of the Model ZB5M implant procedure is the actual error in refractive correction (ie, achieved minus intended), which was less than a mean -0.40 D over the first 2 years of follow-up, increasing to slightly above -0.60 D at the 3 year examination (Fig 3). The proportion of implanted eyes within ±2.00 D of intended correction varied from 88.2% to 98.7% (58.8% to 76.9% within ±1.00 D; 32.4% to 50.0% within ±0.50 D), with the lower percentages being recorded at the final examination. Similar results were also obtained for spherical equivalent refraction, except that both mean values and proportion of eyes within ±0.50, ±1.00, or ±2.00 D of emmetropia were somewhat reduced in comparison to the accuracy refraction data (Fig 1).
As expected, the reduction in spherical equivalent refraction to near emmetropia for the average eye in this study was accompanied by a marked improvement in both near and distance uncorrected visual acuity, increasing from 0.048 and 0.21, respectively, at baseline to approximately 0.5 and 0.7 (Figs 4, 6). Lens implantation resulted in substantial improvement in spectacle-corrected distance visual acuity (but not near visual acuity), with over 90% of treated eyes having a visual acuity score of ^0.5 at each postoperative examination in contrast to only 57.7% at entry (Figs 5, 7). The apparent decline in distance visual acuity and spherical equivalent refraction between years 2 and 3 after Model ZB5M lens implantation is probably attributable to the natural progression of myopia in some eyes, and/or disproportionate representation of eyes having a poorer outcome (implanted during the early learning curve phase of the protocol) in the 3 year data. It is unlikely, however, that the decrease in the values for either efficacy parameter is IOL-related; no evidence exists of an increase in incidence or severity of any complication other than pupil ovalization over the last 2 years of follow-up, and there is no reason to believe that ovalization of the pupil would compromise distance but not near visual acuity, or affect refractive error.
Because the incidence of halos/glare and iris retraction with pupillary ovalization were relatively high in this study, changes have been made in the Model ZB5M lens to redress these treatment-related complications (Figs 9 and 10). The third generation Baikoff MA20 NuVita lens (Bausch and Lomb/Chiron Surgical, Irvine, Calif) has an effective optical diameter increased from 4 to 4.5 mm, with the posterior edge of the optic treated with a patented manufacturing process (Peripheral Detail Technology) to decrease glare. Vaulting has been changed from a step vault to straight angulation to further minimize potential contact between the footplate and the iris, and haptic footplates better conform to iridocorneal angle geometry to disperse compression forces over a broader area.
1. Koch DD. Enter with caution. J Cataract Refract Surg 1996;22:153-154.
2. Baikoff G. Phakic anterior chamber intraocular lenses. Int Ophthalmol Clin 1991;31:75-86.
3. Mimouni F, Colin J, Koffi V, Bonnet P. Damage to the corneal endothelium from anterior chamber intraocular lenses in phakic myopic eyes. Refract Corneal Surg 1991;7:277-281.
4. Saragoussi JJ, Cotinat J, Renard G, Savoldelli M, Abenhaim A, Pouliquen Y. Damage to the corneal endothelium by minus power anterior chamber intraocular lenses. Refract Corneal Surg 1991;7:282-285.
5. Baikoff G, Colin J. Intraocular lenses in phakic patients. Ophthalmol Clin N Am 1992;5:789-795.
6. van der Heijde GL. Some optical aspects of implantation of an IOL in a myopic eye. Eur J Implant Ref Surg 1989;1:245-248
7. Diggle PJ, Liang KY, Zeger SL. Analysis of Longitudinal Data. Oxford, England: Oxford University Press; 1994.
8. Corno F, Ratier C, Simon J, Zanatti C. Noveau refractometre ophthalmologique assiste par microprocesseur. J Fr Ophthalmol 1984;7:485-492.
9. Bourne WM, Kaufman HE. Specular microscopy of human corneal endothelium in vivo. Am J Ophthalmol 1976;81:319-323.
10. Wilson RS, Roper-Hall MJ. Effect of age on the endothelial cell count in the normal eye. Brit J Ophthalmol 1982;66:513-515.
11. Yee RW, Matsuda M, Schultz RO, Edelhauser HF. Changes in the normal corneal endothelial cellular pattern as a function of age. Curr Eye Res 1985;4:671-678.
12. Bourne WM, Nelson LF, Hodge DO. Central corneal endothelial cell changes over a 10-year period. Invest Ophthalmol Vis Sci 1997;38:779-782.
13. Werblin TP. Long-term endothelial cell loss following phacoemulsification: Model for evaluating endothelial damage after intraocular surgery. Refract Corneal Surg 1993;9:29-35.
14. Perez-Santonja JL, Iradier MT, Sanz-Iglesias L, Serrano JM, Zato MA. Endothelial changes in phakic eyes with anterior chamber intraocular lenses to correct high myopia. J Cataract Refract Surg 1996;22:1017-1022.
15. Alio JL, Ruiz-Moreno JM, Artola A. Retinal detachment as a potential hazard in surgical correction of severe myopia with phakic anterior chamber lenses. Am J Ophthalmol 1993;115:145-148.
16. Foss AJE, Rosen PH, Cooling RJ. Retinal detachment following anterior chamber lens implantation for the correction of ultra-high myopia in phakic eyes. Br J Ophthalmol 1993;77:212-213.
17. Leroux les Jardins S, Heligon JP, Özdemir N, Pasquier N, Le Hoang P. Medium-term tolerance of anterior chambr implants in surgical treatment of severe myopia. J France D Ophtal 1995;18:45-49.
Endothelial Cell Density in Eyes Implanted with the ZB5M Phakic IOL*
Complications in 134 Eyes Implanted with the ZB5M IOL